It’s something called ADS (Alternate Data Stream) which you can see as some kind of second hidden file content. Browsers create an ADS with name Zone.Identifier when downloading a file and attach it to the downloaded file. The content of the ADS is the information where the file was downloaded from, i.e. the Zone (3 for Internet) and usually the URL.
Programs and Windows usually use the existence of the Zone.Identifier to show you a warning that a file was downloaded and may pose a risk to your system when opening/exexuting it.
I download a file on Linux, and it has data, a filename, and some permission metadata. That’s it. It sounds like this metadata layer deserves all of the hacks that will come for it.
Most if not all file systems have features that include metadata. For example, ext4 will have a create and modify timestamp, as well as permissions on the file and ownership of the file. Fat also has timestamps but they are less precise. Zfsbootmenu uses zfs metadata to find your key file for encryption.
Your gut reaction of “metadata is over-complication” is not only bizarre (metadata is extremely common in file systems as well as for a variety of file formats - try using musicbrainzpicard without metadata), but makes you look very petulant.
Most file systems have a very limited footprint of metadata. Static information. And they are usually following basic POSIX standards, to ensure that file transfers between mediums are as cross-compatible as possible.
This Alternative Data Stream is now creating this entirely new variable data stream for hackers to hide shit in. No longer can just you scan a file’s data to make sure nothing malicious is in there. Now you need complex logic to be able to both read this new stream, interpret the flags and other metadata, and take all of those different pieces of information and figure out if it’s even worth opening the damn file.
Data is data. Keep data in the data layer. Everything else is secondary, and should be kept tiny.
So you acknowledge metadata is useful but you draw a line because you think it’s too much metadata? Or just the wrong kind, because the POSIX standard is the defining document for your life, a set of rules so perfect that nobody need add functionality past that standard?
It’s something called ADS (Alternate Data Stream) which you can see as some kind of second hidden file content. Browsers create an ADS with name Zone.Identifier when downloading a file and attach it to the downloaded file. The content of the ADS is the information where the file was downloaded from, i.e. the Zone (3 for Internet) and usually the URL.
Programs and Windows usually use the existence of the Zone.Identifier to show you a warning that a file was downloaded and may pose a risk to your system when opening/exexuting it.
What the fuck is this arcane metadata bullshit?
I download a file on Linux, and it has data, a filename, and some permission metadata. That’s it. It sounds like this metadata layer deserves all of the hacks that will come for it.
Most if not all file systems have features that include metadata. For example, ext4 will have a create and modify timestamp, as well as permissions on the file and ownership of the file. Fat also has timestamps but they are less precise. Zfsbootmenu uses zfs metadata to find your key file for encryption.
Your gut reaction of “metadata is over-complication” is not only bizarre (metadata is extremely common in file systems as well as for a variety of file formats - try using musicbrainzpicard without metadata), but makes you look very petulant.
Most file systems have a very limited footprint of metadata. Static information. And they are usually following basic POSIX standards, to ensure that file transfers between mediums are as cross-compatible as possible.
This Alternative Data Stream is now creating this entirely new variable data stream for hackers to hide shit in. No longer can just you scan a file’s data to make sure nothing malicious is in there. Now you need complex logic to be able to both read this new stream, interpret the flags and other metadata, and take all of those different pieces of information and figure out if it’s even worth opening the damn file.
Data is data. Keep data in the data layer. Everything else is secondary, and should be kept tiny.
So you acknowledge metadata is useful but you draw a line because you think it’s too much metadata? Or just the wrong kind, because the POSIX standard is the defining document for your life, a set of rules so perfect that nobody need add functionality past that standard?